Data selection slow for dimension value empty

Dear BW Experts,
I have face the below situation:
    Let say i have dimension A {Date_billing, Date_invoice, Date_reconcilation}, and I have dimension B {Sales Org, Distribution channel}
   Date_billing is always not empty because it is coming from billing document.
   But Date_invoice and Date_reconciliation is sometimes value are empty because those are calculated base on some logic.
When i run querry with selection screen of {Date_billing, Sales org, Distribution Channel}, it runs very fast, it uses star stransformation.
But when i run querry with selection screen of {Date_invoice, Sales org,..}, sometimes it runs very slow and even give dump due to run time long. I have checked and what i found is "If user enters the Date_invoice which is existing in dimension table, then it return value very fast, but if users enters some value of Date_invoice which is not existing in dimension table then i run ver slow and give dump".
To advoid this situation, i have been thinking about NON_STAR_TRANSFORMATION, it is OK. But except this do we have any other solutions, do i missed somthing in my setting.
I have been thinking that I should change selection screen like the way that only date existing in dimension table of Date_invoice will display, but I do not want to do this, what i refer is "Is there any other solutions related to BW selection setting to eliminate those such empty value fastly when inner join happens."
I am looking forward to hearing from you.
Thanks and regards,
KIMC.

Hi BW experts,
   i have checked, the reason is that they use STAR TRANSFORMATION, and we used oracle 10.2.0.2, I really do not know how to calculation cost of querry base on CBO.
   Please guy me, for example, i have a querry, how do i know cost of CBO run if that is STAR TRANSFORMATION and if that is NON STAR TRANSFORMATION.
   I am looking forwards to hearing from you.
thanks and Regards,
KIMC>

Similar Messages

I'm doing a scan around a line by sampling data 360 degrees for every value of z(z is the position on the line). So, that mean I have a double for-loop where I collect the data. The problem comes when I try to plot the data. How should I do?

I'm doing a scan around a line by sampling data 360 degrees for every value of z(z is the position on the line). So, that mean I have a double for-loop where I collect the data. The problem comes when I try to plot the data. How should I do?

Jonas,
I think what you want is a 3D plot of a cylinder. I have attached an example using a parametric 3D plot.
You will probably want to duplicate the points for the first theta value to close the cylinder. I'm not sure what properties of the graph can be manipulated to make it easier to see.
Bruce
Bruce Ammons
Ammons Engineering
Attachments:
Cylinder_Plot_3D.vi ‏76 KB

BPC10 - Data manager package for dimension data export and import

Dear BPC Expers,
Need your help.
I am trying to set up a data manager package for first time to export dimension - master data from one application and import in another application ( both have same properties) .
I created a test data manager package from Organize > add package > with process chain /CPMB/EXPORT_MD_TO_FILE and Add
In the advance tab of each task there are some script logic already populated. please find attached the details of the script logic written under each of the tasks like MD_Source, concvert and target .
I have not done any chnages in the script inside the task .
But when i run the package , I have selected a dimension 'Entity' but in second prompt ,it ask for a transformation file , and syatem autometically add the file ... \ROOT\WEBFOLDERS\COLPAL\FINANCE\DATAMANAGER\TRANSFORMATIONFILES\Import.xls
I have not changed anything there
in the next prmpt , it ask for a output file ..and it won't allow me enter the file name .....i
Not sure how to proceed further.
I shall be greatfull if someone guide me from your experiance how to set up a simple the data manager package for master data export from dimension . Should I update the transformation file in the script for import file and output file in the advance tab. how and what transformation file to be created and link to the data manager package for export / import .
What are the steps to be executed to run the package for exporting master data from dimension and import it another application .
Thanks in advance for your guidance.
Thanks and Regards,
Ramanuj
=====================================================================================================
Detals of the task
Task : APPL_MD-SOURCE
(DIMENSIONMEMBER, %DIMENSIONMEMBERS%, "Please select dimension", "Please select members", %DIMS%)
(TRANSFORMATION,%TRANSFORMATION%,"Transformation file:",,,Import.xls)
(OUTFILE,,"Please enter an output file",Data files (*.txt)|*.txt|All files(*.*)|*.*)
(RADIOBUTTON,%ADDITIONINFO%,"Add other information(Environment,Model,User,Time)?",1,{"Yes","No"},{"1","0"})
(%TEMPNO1%,%INCREASENO%)
(%TEMPNO2%,%INCREASENO%)
(/CPMB/APPL_MD_SOURCE,SELECTION,%DIMENSIONMEMBERS%)
(/CPMB/APPL_MD_SOURCE,OUTPUTNO,%TEMPNO1%)
(/CPMB/EXPORT_MD_CONVERT,INPUTNO,%TEMPNO1%)
(/CPMB/EXPORT_MD_CONVERT,TRANSFORMATIONFILEPATH,%TRANSFORMATION%)
(/CPMB/EXPORT_MD_CONVERT,SUSER,%USER%)
(/CPMB/EXPORT_MD_CONVERT,SAPPSET,%APPSET%)
(/CPMB/EXPORT_MD_CONVERT,SAPP,%APP%)
(/CPMB/EXPORT_MD_CONVERT,OUTPUTNO,%TEMPNO2%)
(/CPMB/FILE_TARGET,INPUTNO,%TEMPNO2%)
(/CPMB/FILE_TARGET,FULLFILENAME,%FILE%))
(/CPMB/FILE_TARGET,ADDITIONALINFO,%ADDITIONINFO%))
Task : EXPORT_MD_CONVERT
(DIMENSIONMEMBER, %DIMENSIONMEMBERS%, "Please select dimension", "Please select members", %DIMS%)
(TRANSFORMATION,%TRANSFORMATION%,"Transformation file:",,,Import.xls)
(OUTFILE,,"Please enter an output file",Data files (*.txt)|*.txt|All files(*.*)|*.*)
(RADIOBUTTON,%ADDITIONINFO%,"Add other information(Environment,Model,User,Time)?",1,{"Yes","No"},{"1","0"})
(%TEMPNO1%,%INCREASENO%)
(%TEMPNO2%,%INCREASENO%)
(/CPMB/APPL_MD_SOURCE,SELECTION,%DIMENSIONMEMBERS%)
(/CPMB/APPL_MD_SOURCE,OUTPUTNO,%TEMPNO1%)
(/CPMB/EXPORT_MD_CONVERT,INPUTNO,%TEMPNO1%)
(/CPMB/EXPORT_MD_CONVERT,TRANSFORMATIONFILEPATH,%TRANSFORMATION%)
(/CPMB/EXPORT_MD_CONVERT,SUSER,%USER%)
(/CPMB/EXPORT_MD_CONVERT,SAPPSET,%APPSET%)
(/CPMB/EXPORT_MD_CONVERT,SAPP,%APP%)
(/CPMB/EXPORT_MD_CONVERT,OUTPUTNO,%TEMPNO2%)
(/CPMB/FILE_TARGET,INPUTNO,%TEMPNO2%)
(/CPMB/FILE_TARGET,FULLFILENAME,%FILE%))
(/CPMB/FILE_TARGET,ADDITIONALINFO,%ADDITIONINFO%))
Task : FILE_TARGET
(DIMENSIONMEMBER, %DIMENSIONMEMBERS%, "Please select dimension", "Please select members", %DIMS%)
(TRANSFORMATION,%TRANSFORMATION%,"Transformation file:",,,Import.xls)
(OUTFILE,,"Please enter an output file",Data files (*.txt)|*.txt|All files(*.*)|*.*)
(RADIOBUTTON,%ADDITIONINFO%,"Add other information(Environment,Model,User,Time)?",1,{"Yes","No"},{"1","0"})
(%TEMPNO1%,%INCREASENO%)
(%TEMPNO2%,%INCREASENO%)
(/CPMB/APPL_MD_SOURCE,SELECTION,%DIMENSIONMEMBERS%)
(/CPMB/APPL_MD_SOURCE,OUTPUTNO,%TEMPNO1%)
(/CPMB/EXPORT_MD_CONVERT,INPUTNO,%TEMPNO1%)
(/CPMB/EXPORT_MD_CONVERT,TRANSFORMATIONFILEPATH,%TRANSFORMATION%)
(/CPMB/EXPORT_MD_CONVERT,SUSER,%USER%)
(/CPMB/EXPORT_MD_CONVERT,SAPPSET,%APPSET%)
(/CPMB/EXPORT_MD_CONVERT,SAPP,%APP%)
(/CPMB/EXPORT_MD_CONVERT,OUTPUTNO,%TEMPNO2%)
(/CPMB/FILE_TARGET,INPUTNO,%TEMPNO2%)
(/CPMB/FILE_TARGET,FULLFILENAME,%FILE%))
(/CPMB/FILE_TARGET,ADDITIONALINFO,%ADDITIONINFO%))
================================================================================

1. Perhaps you want to consider a system copy to a "virtual system" for UAT?
2. Changes in QAS (as with PROD as well) will give you the delta. They should ideally be clean... You need to check the source system.
Another option is to generate the profiles in the target system. But for that your config has to be sqeaky clean and in sync, including very well maintained and sync'ed Su24 data.
Cheers,
Julius

The sme variable in data selection of inpopackage and value in transfer rul

Hello
I would like to use the same variable in data selection of infopackage (1) and as a value in transfer rules (2).
Example :
If I pass 1 I want to see value1 in data selection of infopackage (1) and as a value in transfer rules (2).
If I pass 2 I want to see value2 in data selection of infopackage (1) and as a value in transfer rules (2).
How is possible?

Hi Aleks,
In the transfer routine your code would be something similar to this:
zvar is a variable
select LOW from RSLDPSEL into zvar where LOGDPID = '<Technical name of Infopackage>' and IOBJNM = '<The Info-object whose value you want to use>'.
*Now assign the value
RESULT = zvar.
Hope this helps.
Bye
Dinesh

Select statement for summing values in column?

How to write a select statement for sap standard table which contains data as given below
col1        col2
A     10
A     20
A     30
B     50
B     60
C     70
Need the output of select in internal table as given below
A     60
B     110
C     70
Regards,
Rachel

Thanks for all the valuable input.
Below I have used sum in select statement(check case 2"),but i am not getting required o/p.
Kindly guide me.
case1 :
select aebeln aebelp bpackno bmenge
into corresponding fields of table itab
from ( eslh as a inner join esll as b
on bpackno = apackno )
where a~ebeln = 6000000100
and a~ebelp = 00010
and a~packno = '0000001025'
order by aebeln aebelp b~packno .
O/p for the above .
ebeln                       ebelp         packno                             menge
6000000100     00010     0000001025              6.000
6000000100     00010     0000001025               5.000
6000000100     00010     0000001025              2.000
6000000100     00010     0000001025              1.000
6000000100     00010     0000001025              2.000
case2:
select aebeln aebelp bpackno sum( bmenge )
into corresponding fields of table itab
from ( eslh as a inner join esll as b
on bpackno = apackno )
where a~ebeln = 6000000100
and a~ebelp = 00010
and a~packno = '0000001025'
group by aebeln aebelp b~packno
order by aebeln aebelp b~packno .
6000000100     00010     0000001025     0.000
Edited by: Rachel on Nov 25, 2008 11:45 AM
Edited by: Rachel on Nov 25, 2008 11:46 AM

Date range mapping for dimension

Hi Guys,
I am looking for a way where I can provide DATE as a range for dimension , Any such possibilities anyone is aware of ?? Please suggest
For string or number I have used something like
Country
A - B
C - D
etc
Salary
10000$ - 50000$
51000$ - 99000$
etc
What I want now is something similar for dates
Date
This Week
This Month
etc
Please suggest
Thanks In Advance

The pattern here is to convert the dates to numerics.
For example, your record could have as properties in YYYYMMDD format and be configured for filtering:
p_start_date = 20140325
At run time, you take the date as entered from the UX and convert it to YYYYMMDD format and create your filter for dates between Jan 1, 2014 and March 31, 2014:
Nf=p_start_date|BTWN+20140101+20140331

How can I switch to Longtext in a variable selection box for entered value

I execute a Web Template and enter a value for example a company variable. After this I press the "Check"-Button and get the short text of the company but I want to see the long text.
The InfoObject settings under Business Explorer is:
Display:                                          Text
Text Type:                                       Default
BEx Description:                              Long description
Selection:                                        No Selection Restriction
Query Def. Filter Value Selection:      Values in Master Data Table
Query Execution Filter Val. Selectn: Values in Master Data Table
Filter Value Rep. At Query Exec.:      Selector Box Without Values
The InfoObject is a reference characteristic and the query was created on BEx 3.5.

I changed the text type to long text, but the result is the same.
The only way to get the desired result is to delete the short text and the middle text in the InfoObjects Text Table Properties. But this is not right way.

Data Selection option for Measures only?

Hi All I have Data Source with 3 dimenstion and few measures. I see that We can pick measures to dispaly,, but can i also pick dimenstion to show on horizonatal a-axis ? In screenshot below, I would like to pick one domension for chart A and pick 2nd dimension for chart B.

It's clear, but the selection you are showing above is only to bind measures.
Why not use a filter panel, dimension panel, etc. to filter?

Querying many tables at once from a select statement for specific values

Hi all,
I'm very new to PL/SQL and have a daunting task ahead of me. I'm pretty sure our database normalization scheme is all out of whack, which I dont have the experience to fix yet, but this is the task ahead of me without getting to change or fix our structure:
in the universe has 1000+ tables.
need--> Some of those tables have fields that contain a value that has to be changed.
The only thing I have to go on is the column name will contain 'EMP'... but there is no specific naming convention for the column names.
When I search on:
select * from all_tab_columns where column_name like UPPER('%EMP%')
It returns 750 rows. (So 750 rows = 750 columns in 500 different tables (some tables have 2 or more columns in it))
tables names are random (table1, table2, etc).
columns names are random with EMP in it (column1,column2,etc).
I have 75 possible field values that I need to check each of those 750 columns for.
Maybe this is too messy, but I'm looking at subqueries, joins, arrays, cursors, etc, anything to minimize the amount of work I have to do.
I think this is too much to do a join with, but that seems to be what all the documentation is pointing me to.
can you point me to a simple example of what i could
write?
the logic i see it would be
run the select statement to grab the column names that meet EMP.
one at a time, search each column in the corresponding table for the 75 values.
return the results of rows for each column/table that meet one of the 75 unique values.
I'll have to also update each of the 75 values to (used to be values 1, 2, 3 ... to new values X1, X2, X3, etc.)
Any help at all would be very helpful,
aspiring pl/sql programmer but having a rough time of it

I think you're looking to do something like the following pseudocode:
BEGIN
    FOR tblrec IN (SELECT table_name, column_name
                     FROM all_tab_columns
                    WHERE column_name LIKE '%EMP%')
    LOOP
        FOR rec IN (SELECT ROWID
                      FROM tblrec.table_name
                     WHERE REGEXP_LIKE(tblrec.column_name, '(expr1|expr2|...)')
        LOOP
            DBMS_OUTPUT.PUT_LINE(
                'Found expression on ROWID ' || rec.rowid
                || ' in column ' || tblrec.column_name
                || ' on table ' || tblrec.table_name);
        END LOOP;
    END LOOP;
END;

Select querry for maximum value

Hi,
I want to select fields A B C D from a database table TAB using some selection criteria from selection screen, but the greatest value of A has to be selected for all unique entries of B C and D. Please suggest an appropriate select querry.
PS: I cannot use SELECT and then DELETE ADJACENT DUPLICATES as it will effect performance in some cases.
Regards.

Try using aggregate function MAX with SELECT statement.
MAX Returns the greatest value in the column determined by the database field f for the selected lines. Specifying DISTINCT does not change the result. NULL values are ignored unless all values in a column are NULL values. In this case, the result is NULL .
Have a look at below link:
[Select Clause|http://www.sts.tu-harburg.de/teaching/sap_r3/ABAP4/select_c.htm]
I hope it helps.
Best Regards,
Vibha
Please mark all the helpful answers

How to set the Default values for Info Objects in Data Selection of InfoPac

Hi All,
Flat file Extracion:
How to set the Default values for Info Objects in Data Selection Tab for Info Package
ex: Fiscal Year Variant Info Object having values 'K4' 'Y2' etc in Flat file
Initially default value(not constant) for this info Object value should be 'K4' in Info Package
If I set data selection value for this info object K4 it will retreive records with this selection only? how to handle
Rgds,
CV

Hi,
suppose as your ex. if you are having fiscalyear variant in the dataselection tab then specify K4 in the from column, again the ficalyearvariant row and click on insert duplicate row at the bottom . you will get another row . In that enter Y2 in the from column. now you can extract K4, y2 values .
haritha

Method which is return's all the refinement state for Dimension.

Hi All,
Is there any java method which is return's all the refinement state for Dimension.
Ex:
Brand
Reebok
Scholl
Sebago
Ravel
When i clicked All Brand from front end it will select all 4 Brands data.
Any help will be great appreciation
Thanks,
Sunil N

Not sure what you want to achieve here. If you want to select all the dimension values, then provided the dimension is Multi-select OR you can have N=Id-1+Id-2+Id-3 and it will return all of them. What I'm not sure on is why you would do this, you'll get back precisely the same results (provided every record is tagged with a brand) as if you did nothing, although I guess you might want to allow them to select all then deselect one or more afterwards ... Note you can also use N=<dimension-id>, where <dimension-id> is the id of the actual dimension, and that will return all records that have a value for that dimension, but that won't give you any descriptors (breadcrumbs) other than the dimension name/id.
Michael

Data selection period seems no use in logical DB PCH??

Hi experts,I have a problem when using LDB PCH.
There are object selection period fields PCHOBEG and PCHOEND.
there are also data selection period fields PCHBEGDA and PCHENDDA .
I set the data selection fields PCHBEGDA & PCHENDDA values at Start-of-selection event.
But this seems to have no effect on the data selection.
on GET OBJEC event,all of the infotype data will be got,but not corrosponding to the data selection period.
Is this normal?or something wrong?Thanks.

In PCH you have two periods. One for selecting objects and one for selecting data (infotypes).
LDB PCH works like PNP or PNPCE

Can't find 0FYTLFP OLAP Variable in InfoPackage Data Selection tab on 3.5

On BW 3.5, when generating an InfoPackage with the datasource as Purchasing Data(80PUR_C01), under Data Selection tab, for the InfoObject 0FISCPER, I selected 7 (OLAP variables)for the Type column, then click the button next to it and under the column of Details for Type to bring up a window, but when expending the OLAP Variable field list, can't find the following value in the list which shows up on BW 3.1:
0FYTLFP     0FISCPER Cumulated to Last Fiscal Year/Period (SAP Exit)
How to make the above value show up in the OLAP Variable list?   I thought maybe the variable 0FYTLFP needs to be installed in Business Contents, but there is no variable object type in Business Content. It sounds like a SAP Exit, but how to make the SAP Exit value show up on BW 3.5?
Thanks for everyone's input!

hey Rohit,
Besides the three variable values in the list on BW 3.5, now I find a new one in the list, called "0E_ROLE     0SR_ROLE     Role (SAP Exit)", I have no idea why this one shows up now. Now all the four variables in the list are:
1. 0CML12LM     0CALMONTH     Last 12 Months Excluding the Current Month
2. 0CMONTH     0CALMONTH     Current Calendar Month
3. 0E_ROLE     0SR_ROLE     Role (SAP Exit)
4. 0S_RQMRC     0REQUID     Most Current Data
0FYTLFP is still not in the list!   We can't try to create BEx query since our BW 3.5 system is just installed and we have not got to the step to ceate a query. I try the 2nd way you suggested by pressing F6 on this variable with the typed in value 0FYTLFP, get the following msg "BEx variable 0FYTLFP could not be evaluated - No error message", that means it would not work on BEx query as well, right?
What could be the reason that we can't see 0FYTLFP variable in our new installed BW 3.5? Do you think our BASIS missed installing sth?
Thanks
Message was edited by: Kevin Smith

BW : Data selection customizing in InfoPackage

Hi everybody!
In BW, RSA1, I have an infopackage, that I run monthly.
It's not a delta.
I have a field "month" in the data selection tab.
This value should be the same as the loading month.
Is there a way to fill the field with an external routine ?
Thanks a lot for any reply.
François Leyssenne

Thanks a lot!
I never saw that stuff, yet it seems very very usefull!
François

Data selection slow for dimension value empty

Similar Messages

Maybe you are looking for